Algorithm 1 DBSCAN: Density-based spatial clustering of applications with noise

Require:$|{\mathit{H}}_{N\times M}|$: amplitude of $\mathit{H}$; $MinPts$: minimum neighbor number requirement for a central point of a cluster; $Eps$: neighborhood radius;

Ensure: clustering result C

1: $Visi{t}_{N\times 1}={[0,0,\dots ,.0]}^T$: mark all points in $\left|\mathit{H}\right|$ as unvisited points, $Nois{e}_{N\times 1}={[0,0,\dots ,.0]}^T$: mark all points in $\left|\mathit{H}\right|$ as non-noise points, $ID{X}_{N\times 1}={[0,0,\dots ,.0]}^T$: mark all points in $\left|\mathit{H}\right|$ as the state of not adding any clusters, $C=0$: initial number of cluster;

2: Normalize $\left|H\right|$

3: for each point p in $\left|\mathit{H}\right|$ do

4:   if $Visit\left(p\right)=0$ then

5:    $Visit\left(p\right)=1$;

6:    Calculate the Euclidean distance between this point and the other points and get a set of neighbors $N1$ which have a distance of less than $Eps$;

7:    if $num\left(N1\right)<MinPts$ then

8:     $Noise\left(p\right)=1$;

9:    else

10:     $C=C+1$;

11:     $k=1$;

12:     repeat

13:      $k=k+1$; $IDX\left(k\right)=C$;

14:      if $Visit\left(k\right)=0$ then

15:       $Visit\left(k\right)=1$;

16:       Calculate the Euclidean distance between this point and the other points and get a set of neighbors $N2$ which have a distance of less than $Eps$;

17:       if $num\left(N2\right)>=MinPts$ then

18:        $N1=N1\cup N2;$ ;

19:       end if

20:      end if

21:      if $IDX\left(k\right)=0$ then

22:       $IDX\left(k\right)=C$

23:      end if

24:     until k>num(N1)

25:    end if

26:   end if

27: end for